Matt Ridley

where there is less certainty, such as the APOE case, the test would be of still less value. You can still - if you are very lucky — have two E4 genes and live to an old age with no symptoms, just as you can still - if you are very unlucky - have no E4 genes and get Alzheimer's at sixty-five. Since a diagnosis of two E4 genes is neither sufficient nor necessary to predict Alzheimer's, and since there is no cure, you should not be offered the test unless you are already symptomatic.

At first I found all these arguments convincing, but now I am not so sure. After all, it has been considered ethical to offer people the test for the H I V virus if they want it, even though A I D S was (until recently) incurable. A I D S is not an inevitable outcome of H I V infection: some people survive indefinitely with H I V infection. True, there is in the case of A I D S the additional interest of society in preventing the spread of the infection, which does not apply to Alzheimer's disease, but it is the individual at risk we are considering here, not society at large. The Nuffield Council addresses this argument by implicitly making a distinction between genetic and other tests. To attribute a person's susceptibility to an illness to their genetic make-up distorts attitudes, argued the report's author, Dame Fiona Caldicott. It makes people believe wrongly that genetic influences are paramount and causes them to neglect social and other causes; that, in turn, increases the stigma attached to mental illness.8

This is a fair argument unfairly applied. The Nuffield Council is operating a double standard. 'Social' explanations of mental problems offered by psychoanalysts and psychiatrists are licensed to practise on the flimsiest of evidence, yet they are just as likely to stigmatise people as genetic ones. They continue to flourish while the great and the good of bioethics outlaw diagnoses that are supported by evidence merely because they are genetic explanations. In striving to find reasons to outlaw genetic explanations while allowing social ones to flourish, the Nuffield Council even resorted to calling the predictive power of the APOE4 test Very low' - bizarre word-ing for an eleven-fold difference in risk between the E4/E4S and 2 6 6 G E N O M E

the E3/E3s.9 As John Maddox has commented,10 citing APOE as a case in point, 'There are grounds for suspecting that physicians are not pursuing valuable opportunities out of diffidence at revealing unwelcome genetic information to their patients . . . but diffidence can be taken too far.'

Besides, although Alzheimer's disease is incurable, there are already drugs that alleviate some of the symptoms and there may be precautions of uncertain value that people can take to head it off. Is it not better to know if one should take every precaution?

If I had two E4 genes, I might well want to know so that I could volunteer for trials of experimental drugs. For those who indulge in activities that raise their risk of Alzheimer's disease, the test certainly makes sense. It is, for example, now apparent that professional boxers who have two E4 genes are at such risk of developing early Alzheimer's that boxers are indeed best advised to take a test and not box if they find themselves with two E4S. One in six boxers get Parkinson's disease or Alzheimer's — the microscopic symptoms are similar, though the genes involved are not - by the age of fifty, and many, including Mohammed Ali, suffer even younger. Among those boxers who do get Alzheimer's, the E4 gene is unusually common, as it is among people who suffer head injury and later turn out to have plaques in their neurons.

What is true for boxers may be true for other sports in which the head is struck. Alerted by anecdotal evidence that many great footballers sink into premature senility in old age — Danny Blanchflower, Joe Mercer and Bill Paisley being sad, recent examples from British clubs — neurologists have begun to study the prevalence of Alzheimer's disease in such sportsmen. Somebody has calculated that a soccer player on average heads the ball 800 times in a season; the wear and tear could be considerable. A Dutch study did indeed find worse memory loss in footballers than in other sportsmen and a Norwegian one found evidence of brain damage in soccer players.

Once more it is plausible that the E4/E4 homozygotes might benefit from at least knowing at the outset of their careers that they were specially at risk. As somebody who frequently hits his head on door P R E V E N T I O N 2 6 7

frames because architects have not made them big enough for tall people to walk through, I wonder myself what my APOE genes looks like. Maybe I should have them tested.

Testing could be valuable in other ways. At least three new Alzheimer's drugs are in development and testing. One that is already here, tacrine, is now known to work better in those with E3 and E2 genes than in E4 carriers. Again and again the genome drives home the lesson of our individuality. The diversity of humanity is its greatest message. Yet there is still a marked reluctance in the medical profession to treat the individual rather than the population.

A treatment that is suitable for one person may not suit another.

Dietary advice that could save one person's life might do no good at all to another. The day will come when a doctor will not prescribe you many kinds of medicine until he has checked which version of a gene or genes you have. The technology is already being developed, by a small Californian company called Affymetrix among others, to put a whole genome-full of genetic sequences on a single silicon chip. One day we might each carry with us exactly such a chip from which the doctor's computer can read any gene the better to tailor his prescription to us.11

Perhaps you have already sensed what the problem with this would be — and what is the real reason behind the experts' squeam¬

ishness about APOE tests. Suppose I do have E4/E4 and I am a professional boxer. I therefore stand a much higher than average chance of contracting angina and premature Alzheimer's disease.

Suppose that today, instead of going to see my doctor, I am going to see an insurance broker to arrange a new life-insurance policy to go with my mortgage, or to get health insurance to cover future illness. I am handed a form and asked to fill in questions about whether I smoke, how much I drink, whether I have A I D S and what I weigh. Do I have a family history of heart disease? - a genetic question. Each question is designed to narrow me down into a particular category of risk so that I can be quoted an appropriately profitable, but still competitive premium. It is only logical that the insurance company will soon ask to see my genes as well, to ask if 268 G E N O M E

I am E4/E4, or if I have a pair of E3s instead. Not only does it fear that I might be loading up on life insurance precisely because I know from a recent genetic test that I am doomed, thus ripping it off as surely as a man who insures a building he plans to burn down. It also sees that it can attract profitable business by offering discounts to people whose tests prove reassuring. This is known as cherry picking, and it is exactly why a young, slim, heterosexual non-smoker already finds he can get life insurance cheaper than an old, plump, homosexual smoker. Having two E4 genes is not so very different.

Little wonder that in America health-insurance companies are already showing interest in genetic tests for Alzheimer's, a disease that can be very costly for them (in Britain, where health cover is basically free, the main concern is life insurance). But mindful of the fury the industry unleashed when it began charging homosexual men